Hyperintense vessels on FLAIR: a useful non-invasive method for assessing intracerebral collaterals

Eur J Radiol. 2011 Dec;80(3):786-91. doi: 10.1016/j.ejrad.2010.09.043. Epub 2010 Nov 9.


Objective: This study was aimed to evaluate relationship between hyperintense vessels (HV) on fluid-attenuated inversion recovery (FLAIR) and artery steno-occlusion related intracerebral collaterals.

Materials and methods: A total of 233 patients with 260 atherosclerotic lesions in the M1 segment of the middle cerebral artery (MCA) were examined with FLAIR and digital subtraction angiography (DSA). HV were graded as 0, 1, 2 and 3 by its distributions in the MCA territory. Grade 0 indicated no HV; Grade 1 indicated the HV limited in Sylvian fissure; Grade 2 indicated the HV limited in Sylvian fissure and the temporal-occipital junction; Grade 3 indicated the HV extended to frontal-parietal lobes. Collateral blood flows were classified by DSA results. The relationship between HV grades and patterns of collateral flows was analyzed.

Results: HV were observed in 76 out of 260 hemispheres. For patients with Grade 1 HV, most of their collateral flows (80.8%) were antegrade; for patients with Grade 2, the retrograde leptomeningeal flows were commonly manifested as anterior cerebral artery to MCA (75%); for patients with Grade 3 HV, most of the retrograde leptomeningeal flows were manifested as posterior cerebral artery to MCA (81.8%). As the grade HV increased, the frequency of retrograde leptomeningeal collateral from ACA to MCA decreased (100% to 75% and to 18.2%), and increased (0% to 25% and to 81.8%) for the retrograde leptomeningeal collateral via PCA to MCA (P<0.001).

Conclusions: The HV could assess non-invasively intracerebral collaterals in patients with steno-occlusive lesions of M1 segment of MCA.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Collateral Circulation*
  • Humans
  • Infarction, Middle Cerebral Artery / pathology*
  • Magnetic Resonance Angiography / methods*
  • Male
  • Middle Aged
  • Models, Biological
  • Models, Statistical
  • Reproducibility of Results
  • Sensitivity and Specificity